4chan-gpt

Trained on a corpus of 4chan /pol/ and /b/ data collected by SilverFlag. This is a fine tune of OpenAI's open source gpt-2 architecture. SilverFlag takes no responsibility for content inputted or generated from this model. Use with caution.


Controls randomness. Higher values make output more creative and random, lower values make it more focused and deterministic.
Nucleus sampling. Lower values mean only the most likely words are considered. Higher values allow more diverse vocabulary.
The maximum length of the generated response.